The world's first Crowd Behavior Analysis Technology that detects congestion status and anomalies from crowd footage!
- MiyanoIt is difficult for the human eye to notice from security camera footage when someone falls over in a crowded environment. But Crowd Behavior Analysis Technology enables anomalies to be detected immediately by picking up on changes in the behavior of people in the area. In addition to situations where people are stopped or surrounding someone like the video I just showed you, tracking changes in crowd status in real time also enables the quick detection of abnormal crowding when a rush of people leads to an extremely high crowd density, such as on train station platforms and at the entrance to event venues.
- MitaI see.
- MiyanoFor another example, if a knife-wielding man was on the rampage at an airport, etc., everyone in the area would flee, right? We can also detect groups of people running away based on the flow of people, so surveillance staff would be able to rush to that spot quickly and minimize the damage caused. This technology for detecting crowded environments or anomalies by analyzing crowd footage is a world first.
- MitaI feel like this technology will offer a lot of peace of mind!
- MiyanoThat's kind of you to say. Incidentally, you can change the threshold value based on the location or purpose. For example, in places where people don't stop often, you can set an anomaly to be detected when someone stops even for a brief moment. Alternatively, in a plaza or other area where people stop often, you can set the system to ignore people stopped for longer periods of time. This lets you tailor values to the on-site situation.
- MitaThat means you can use it for all sorts of things!
Big data enables the prediction of congestion and danger!
- MiyanoAnother thing that the use of Crowd Behavior Analysis Technology can enable is the prediction of congestion or danger.
- MitaHuh? You can predict where will be crowded, or where something dangerous will happen?
- MiyanoWe can. It all comes down to so-called "big data." By analyzing large volumes of data obtained through Crowd Behavior Analysis Technology, it is possible to predict congestion. For example, knowing which ticket gates and bus stops will get extremely crowded when a train accident occurs on a Friday evening enables you to place security guards accordingly.
- MitaThat sounds really helpful. It would be so much better if something could be done about transportation for fireworks displays, and the routes you take to venues, etc. By the way, can you use Crowd Behavior Analysis Technology anywhere there's a security camera?
- MiyanoYes. That's what we anticipate. Special cameras installed on the ceiling that point straight down can detect numbers of people, but are limited in terms of the range they can shoot and the purposes for which they can be used. There is also the cost associated with installing new cameras. One of the strengths of our technology is the fact analysis is possible using cameras that are already installed, instead of requiring new ones.
- MitaThat really lowers the bar of entry when implementing it! I get the feeling that it has huge potential. Thank you for sparing the time to speak with me today.
- MiyanoThank you for hearing me out. It was a pleasure.
In this installment, we heard about the Crowd Behavior Analysis Technology that enables crowd footage showing many people to be analyzed. I was surprised to find that it can detect congestion status in such detail, but it was really eye-opening when I learned it could detect crowd "anomalies" that occur due to incidents or accidents! I hope to see this technology put to use in a variety of places, keeping us safe in our daily lives. See you in the next installment of "MiTA TV"!
(Published June 17, 2014)